FACIL
FACIL copied to clipboard
How to integrate ViT in Networks?
I want to use Huggingface's ViTForImageClassification. How do I integrate it in FACIL? I want to load the pretrained model 'google/vit-base-patch16-224'. I have read the instructions to add networks in readme of networks. However I am still not sure how to implement it. How do I set "self.head_var = 'fc'" when head is changed by "model.classifier = nn.Linear(768, num_classes)"? How exactly will a class even be created in this case?